Working Group I: The Scientific Basis

Other reports in this collection

7.6.5 El Niño-Southern Oscillation (ENSO)

Figure 7.8:
Simplified principal feedback loops active in El Niño- Southern Oscillation (ENSO). The fast loop (right) gives rise to an instability responsible for the development of an El Niño, the slow loop (left) tends to dampen and reverse the anomalies, so that together, these processes excite oscillations.

The strongest natural fluctuation of climate on interannual time-scales is the El Niño-Southern Oscillation (ENSO) phenomenon, and ENSO-like fluctuations also dominate decadal time-scales (sometimes referred to as the Pacific decadal oscillation). ENSO originates in the tropical Pacific but affects climate conditions globally. The importance of changes in ENSO as the climate changes and its potential role in possible abrupt shifts have only recently been appreciated. Observations and modelling of ENSO are addressed in Chapters 2, 8 and 9; here the underlying processes are discussed. Observational and modelling results suggest that more frequent or stronger ENSO events are possible in the future. Because social and ecological systems are particularly vulnerable to rapid changes in climate, for the next decades, these may prove of greater consequence than a gradual rise in mean temperature. ENSO processes

ENSO is generated by ocean-atmosphere interactions internal to the tropical Pacific and overlying atmosphere. Positive temperature anomalies in the eastern equatorial Pacific (characteristic of an El Niño event) reduce the normally large sea surface temperature difference across the tropical Pacific. As a consequence, the trade winds weaken, the Southern Oscillation index (defined as the sea level pressure difference between Tahiti and Darwin) becomes anomalously negative, and sea level falls in the west and rises in the east by as much as 25 cm as warm waters extend eastward along the equator. At the same time, these weakened trades reduce the upwelling of cold water in the eastern equatorial Pacific, thereby strengthening the initial positive temperature anomaly. The weakened trades also cause negative off-equatorial thermocline depth anomalies in the central and western Pacific. These anomalies propagate westward to Indonesia, where they are reflected and propagate eastward along the equator. Thus some time after their generation, these negative anomalies cause the temperature anomaly in the east to decrease and change sign. The combination of the tropical air-sea instability and the delayed negative feedback due to sub-surface ocean dynamics can give rise to oscillations (for a summary of theories see Neelin et al., 1998). Two of these feedbacks are schematically illustrated in Figure 7.8. Beyond influencing tropical climate, ENSO seems to have a global influence: during and following El Niño, the global mean surface temperature increases as the ocean transfers heat to the atmosphere (Sun and Trenberth, 1998).

Box 7.2: Changes in natural modes of the climate system.

Observed changes in climate over the Northern Hemisphere in winter reveal large warming over the main continental areas and cooling over the North Pacific and North Atlantic. This “cold ocean – warm land” pattern has been shown to be linked to changes in the atmospheric circulation, and, in particular, to the tendency in the past few decades for the North Atlantic Oscillation (NAO) to be in its positive phase. Similarly, the Pacific-North American (PNA) teleconnection pattern has been in a positive phase in association with a negative Southern Oscillation index or, equivalently, the tendency for El Niño-Southern Oscillation (ENSO) to prefer the warm El Niño phase following the 1976 climate shift (Chapter 2). Because of the differing heat capacities of land and ocean, the “cold ocean-warm land” pattern has amplified the Northern Hemisphere warming. A fingerprint of global warming from climate models run with increasing greenhouse gases indicates greater temperature increases over land than over the oceans, mainly from thermodynamic (heat capacity and moisture) effects. This anthropogenic signal is therefore very similar to that observed, although an in-depth analysis of the processes involved shows that the dynamical effects from atmospheric circulation changes are also important. In other words, the detection of the anthropogenic signal is potentially masked or modified by the nature of the observed circulation changes, at least in the northern winter season. The detection question can be better resolved if other seasons are also analysed (Chapter 12). Attribution of the cause of the observed changes requires improved understanding of the origin of the changes in atmospheric circulation. In particular, are the observed changes in ENSO and the NAO (and other modes) perhaps a consequence of global warming itself?

There is no simple answer to this question at present. Because the natural response of the atmosphere to warming (or indeed to any forcing) is to change large-scale waves, some regions will warm while others cool more than the hemispheric average, and counterintuitive changes can be experienced locally. Indeed, there are preferred modes of behaviour of the atmospheric circulation, sometimes manifested as preferred teleconnection patterns (see this chapter) that arise from the planetary waves in the atmosphere and the distribution of land, high topography, and ocean. Often these modes are demonstrably natural modes of either the atmosphere alone or the coupled atmosphere-ocean system. As such, it is also natural for modest changes in atmospheric forcing to project onto changes in these modes, through changes in their frequency and preferred sign, and the evidence suggests that changes can occur fairly abruptly. This is consistent with known behaviour of non-linear systems, where a slow change in forcing or internal mechanisms may not evoke much change in behaviour until some threshold is crossed at which time an abrupt switch occurs. The best known example is the evidence for a series of abrupt climate changes in the palaeoclimate record apparently partly in response to slow changes in sea level and the orbit of the Earth around the Sun (Milankovitch changes, see Chapter 2). There is increasing evidence that the observed changes in the NAO may well be, at least in part, a response of the system to observed changes in sea surface temperatures, and there are some indications that the warming of tropical oceans is a key part of this (see this chapter for more detail). ENSO is not simulated well enough in global climate models to have confidence in projected changes with global warming (Chapter 8). It is likely that changes in ENSO will occur, but their nature, how large and rapid they will be, and their implications for regional climate change around the world are quite uncertain and vary from model to model (see this chapter and Chapter 9). On time-scales of centuries, the continuing increase of greenhouse gases in the atmosphere may cause the climate system to cross a threshold associated with the Atlantic thermohaline circulation: beyond this threshold a permanent shut-down of the thermohaline circulation results (see this chapter and Chapter 9).

Therefore, climate change may manifest itself both as shifting means as well as changing preference of specific regimes, as evidenced by the observed trend toward positive values for the last 30 years in the NAO index and the climate “shift” in the tropical Pacific about 1976. While coupled models simulate features of observed natural climate variability such as the NAO and ENSO, suggesting that many of the relevant processes are included in the models, further progress is needed to depict these natural modes accurately. Moreover, because ENSO and NAO are key determinants of regional climate change, and they can possibly result in abrupt changes, there has been an increase in uncertainty in those aspects of climate change that critically depend on regional changes.

Figure 7.9: Darwin Southern Oscillation Index (SOI) represented as monthly surface pressure anomalies in hPa. Data cover the period from January 1882 to December 1998. Base period climatology computed from the period January 1882 to December 1981. The step function fit is illustrative only, to highlight a possible shift around 1976 to 1977.

The shifts in the location of the organised rainfall in the tropics and the associated latent heat release alters the heating patterns of the atmosphere which forces large-scale waves in the atmosphere. These establish teleconnections, especially the PNA and the southern equivalent, the Pacific South American (PSA) pattern, that extend into mid-latitudes altering the winds and changing the jet stream and storm tracks (Trenberth et al., 1998), with ramifications for weather patterns and societal impacts around the world.

Another related feedback occurs in the sub-tropics. The normally cold waters off the western coasts of continents (such as California and Peru) encourage the development of extensive low stratocumulus cloud decks which block the Sun, and this helps keep the ocean cold. A warming of the waters, such as during El Niño, eliminates the cloud deck and leads to further sea surface warming through solar radiation. Kitoh et al. (1999) found that this mechanism could lead to interannual variations in the Pacific Ocean without involving equatorial ocean dynamics. Currently, stratocumulus decks are not well simulated in coupled models, resulting in significant deviations of SST from the observed (see Chapter 8, Figure 8.1).

Indices of ENSO for the past 120 years (Figure 7.9), indicate that there is considerable variability in the ENSO cycle in the modern record. This variability has been variously attributed to: (i) stochastic forcing due to weather and other high-frequency “noise”, and the Madden-Julian intra-seasonal oscillation in particular; (ii) deterministic chaos arising from internal non-linearities of the tropical Pacific ENSO system; (iii) forcing within the climate system but external to the tropical Pacific, and (iv) changes in exogenous forcing (see Neelin et al., 1998 and references therein). Palaeo-proxies, archaeological evidence, and instrumental data (see Chapter 2) all indicate variations in ENSO behaviour over the past centuries, and throughout the Holocene. Much of this variability appears to be internal to the Earth’s climate system, but there is evidence that the rather weak forcing due to orbital variations may be responsible for a systematic change to weaker ENSO cycles in the mid-Holocene (Sandweiss et al., 1996; Clement et al., 1999; Rodbell et al., 1999). However, it appears that the character of ENSO can change on a much faster time-scale than that of small amplitude insolation change imposed by the Earth’s varying orbit. The inference to be drawn from observed ENSO variability is that small forcings are able to cause large alterations in the behaviour of this non-linear system.

Continues on next page

Other reports in this collection

IPCC Homepage